Objective: The study aimed at evaluating the activity and toxicity of gemcitabine monochemotherapy in a unselected series of elderly patients with advanced bladder cancer. The secondary objectives were to establish whether there is a correlation between treatment and Comprehensive Geriatric Assessment (CGA) and, in addition, to determine overall patient survival. Methods: Treatment consisted of six courses of chemotherapy with gemcitabine at a dosage of 1,200 mg/m<sup>2</sup> on days 1 and 8, every 21 days. CGA, as described by Gruppo Italiano di Oncologia Geriatrica, was assessed for evaluating the functional status of patients before, during, and after treatment. Results: Twenty-five patients were enrolled (M/F 22/3), 22 of these were evaluable for response and 23 for toxicity. Characteristics of patients: median age 76 years (range 71–87); ECOG performance status (PS) 1 in 12 patients and 2 in 13 patients; clinical stage III in 6 patients and IV in 19 patients. At the end of the therapy the parameters of CGA improved in 4 cases (17%), remained unchanged in 17 cases (74%) and worsened only in 2 cases (9%). Two patients were not evaluable. Response evaluation showed 3 (13.5%) complete responses (CRs) and 7 (32%) partial responses (PRs), for an overall response rate of 45.5% [95% confidence interval (CI), 24.3–65.7%]. Three (13.5%) patients had stable disease (SD ) and 9 (41%) disease progression (DP). Median overall survival was 8 months and median time to progression was 5 months. Treatment was generally well tolerated, with 1 patient having grade 3 gastrointestinal toxicity and 3 having grade 4 neutropenia. Conclusions: We conclude that gemcitabine can be safely administered in monochemotherapy, is effective and does not worsen the functional status of elderly patients with advanced bladder cancer.